IonQ Forte run beats classical algorithm on a non-abelian hidden subgroup task
Oliver O'Brien of the University of Cambridge and colleagues ran a hidden subgroup problem for q-hedral groups on the IonQ Forte trapped-ion computer and compared four measurement bases: weak Fourier sampling, strong Fourier sampling in an adapted basis, strong sampling of adjacent pairs, and forgetful abelian sampling. For p = 7 and q = 6, the hardware exceeded the success probability of an exhaustively optimised classical algorithm for up to three oracle queries. Weak sampling revealed nothing about the hidden subgroup, the adapted basis worked best for large q, adjacent pairs did better for small q, and forgetful abelian sampling plateaued at a success probability of (p + q-1)/(pq). The team ran 62 circuits of 2,048 shots each between October 9 and 12, 2026, with gate-folded copies bringing the total to 153,600 shots, and limited tests to p = 3, 5 and 7 to keep noise manageable. The work is an arXiv preprint (not yet peer reviewed), and the data, code and circuits are publicly available.